US5596687A - Apparatus and method for addressing pixel values within an image pyramid using a recursive technique - Google Patents
Apparatus and method for addressing pixel values within an image pyramid using a recursive technique Download PDFInfo
- Publication number
- US5596687A US5596687A US08/282,296 US28229694A US5596687A US 5596687 A US5596687 A US 5596687A US 28229694 A US28229694 A US 28229694A US 5596687 A US5596687 A US 5596687A
- Authority
- US
- United States
- Prior art keywords
- value
- memory
- address
- pixel
- values
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
- 238000000034 method Methods 0.000 title claims abstract description 22
- 230000004044 response Effects 0.000 claims 4
- 238000002059 diagnostic imaging Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 6
- 238000013507 mapping Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 101100386054 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) CYS3 gene Proteins 0.000 description 1
- 239000011449 brick Substances 0.000 description 1
- 101150035983 str1 gene Proteins 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4007—Scaling of whole images or parts thereof, e.g. expanding or contracting based on interpolation, e.g. bilinear interpolation
Definitions
- the invention relates to address generators for accessing memory and, more particularly, to methods and apparatus for generating addresses to access elements or pixel values of an image pyramid stored in memory.
- three-dimensional objects are often defined as a set of polygons, which are projected to a two-dimensional screen space, e.g., a screen of a computer monitor, to give the illusion of depicting a three-dimensional object therein.
- the polygons are "covered” with a texture that is scaled to the shape of the polygon.
- the operation that combines a polygon with a particular texture is known in the art as texture mapping.
- a texture mapping operation combines a scaled and transformed polygon with a similarly scaled and transformed texture to achieve an illusion of three-dimensionality within a two-dimensional screen space.
- Textures are typically stored in memory within the graphics system as a set of identical images each having a different level of resolution. Such an arrangement of textures is known as an image pyramid. The resolution of successive levels of a pyramid differ by a power of two such that each successive level of the pyramid has one-quarter of the pixels contained in the previous (underlying) level.
- each texture is represented by a plurality of color pyramids, e.g., a red pyramid, a green pyramid and a blue pyramid.
- the graphics system produces three memory addresses and accesses three memory locations.
- U and V representing Cartesian coordinates into a base image, i.e., the base of the image pyramid
- D representing a pyramid level.
- U, V, and D define a location of a particular pixel value within a particular image pyramid, e.g., a red pyramid.
- a pixel value necessary to produce an accurate output image may have coordinates that lie between actual pixel locations in a given level of a pyramid. Furthermore, the necessary pixel value may require a resolution that lies between levels. Consequently, the texture mapping function interpolates the nearest pixel values to determine the unknown pixel value. Specifically, for pixel values that are defined between levels of the pyramid, four pixel values in a level above the target pixel location and four pixel values below the target pixel location are tri-linearly interpolated to produce the target pixel value for the target location in an output image.
- the present invention overcomes the disadvantages heretofore associated with the prior art by selecting pixel values from an image pyramid using relatively few instruction cycles.
- the invention is apparatus and a concomitant method for accessing an image pyramid that is sequentially stored in a memory.
- the invention uses an integer portion of the standard U, V, and D values that define a target pixel location within an image pyramid to determine a first address of a pixel value near the target location within the memory. From this first address, the invention determines another seven addresses. These eight addresses are used to recall pixel values that are proximate the target pixel location. These eight pixel values can then be used in a tri-linear interpolation to determine the target pixel value.
- the invention temporarily stores the integer portions of the U, V, and D values in registers. These values are used to determine initial values for an offset register, a base register, and a stride register. Each memory address is defined by combining the offset register value with a base register value.
- the base register value defines the location of the first memory location used for a particular image pyramid. All other values in that pyramid sequentially follow from that address.
- the offset register value determines how far from that base value a particular pixel is located in memory. As such, by updating the offset register value, any location in the pyramid can be addressed.
- the offset register value is periodically updated using the stride register value or by adding or subtracting one to/from the present offset register value. Consequently, from a given initial offset value, adding one addresses the next sequential pixel value, subtracting one addresses the previous pixel value, adding the stride value addresses a pixel in a next column of pixel values, adding the stride register value squared addresses a pixel value in the next level of the image pyramid, and so on.
- the invention appropriately uses the offset register value and the stride register value to efficiently generate addresses for eight pixel values in a particular image pyramid. These eight pixel values can then be used to compute the target pixel value.
- FIG. 1 shows a representation of eight pixel locations within two levels in an image pyramid that are used to compute a target pixel value located between the two levels;
- FIG. 2 depicts a block diagram of hardware used to determine eight memory addresses for pixel values to be used in computing a target pixel value located between levels in an image pyramid;
- FIGS. 3A and B taken together, depict a flow chart of a routine illustrating the operation of the hardware of FIG. 2.
- FIG. 1 depicts a portion 100 of an image pyramid.
- This portion 100 is a portion of an image pyramid that contains a plurality of layers, e.g., five, having a number of pixels in each layer, e.g., 128 pixels in layer 1, 64 pixels in layer 2, 16 pixels in layer 3 and so on.
- the depicted portion contains eight pixel locations located in two levels, i.e., four pixel locations per level. To differentiate the levels, the first level (D i ) uses X,Y coordinates and the second level (D i+1 ) uses A,B coordinates to designates the respective locations of pixels in the respective levels. These eight pixel locations are the eight locations that are nearest a target pixel location 102.
- the present invention determines these eight pixel locations as memory addresses within a memory that sequentially stores all the pixel values in the pyramid.
- the eight pixel values located at the eight pixel memory addresses are used as interpolants in a tri-linear interpolation to determine a pixel value at the target location 102 between the known values.
- image pyramids are addressed using three coordinates U, V, and D.
- U and V represent the two dimensional address of a base coordinate (corner) of the pyramid and D designates a particular level.
- These coordinates are respectively comprised of an integer portion (i) and a fractional portion (f).
- U can be expressed as U i +U f
- V can be expressed as V i +V f
- D can be expressed as D i +D f .
- these coordinates determine a location within the pyramid of a target pixel location 102 and a pixel value for that location.
- the present invention utilizes the integer portion of these coordinates to determine eight pixel locations and values to be used in computing a pixel value for the pixel location 102 located between the known pixel locations.
- the fractional portion of each coordinate is used in performing the tri-linear interpolation used to compute the pixel value.
- the interpolation process is a conventional operation that does not form any part of the present invention.
- TID texture identification
- U and V values are stored in memory such that the system can access the table whenever a certain texture is requested and determine the U and V coordinates for the desired texture.
- the D coordinate is determined from the desired resolution at that location in the output image.
- the present invention rapidly determines eight memory addresses that contain pixel values that are used to generate a pixel value for an output image.
- the image pyramid to be addressed contains RGB data at each memory location, e.g., a 32 bit word.
- the invention produces eight memory locations to enable the system to compute a pixel value.
- the texture data is divided amongst three (or more) pyramids, e.g., one pyramid each for red, blue and green, then the inventive address generator shall be operated thrice to produce a memory address for each pyramid. As such, the interpolation will need 24 memory locations to retrieve 24 pixel values (eight for each color).
- the following disclosure also assumes that the image pyramid is stored in a word serial manner.
- the base pixel value (32-bit word) of the pyramid is stored at a particular location and the remaining pixel values for a given pyramid are sequentially stored in memory locations from the base location.
- FIG. 2 depicts a block diagram of the inventive memory address generator 200.
- the generator can be implemented using either discrete components or within a microprocessor.
- the generator has as its input the values TID, U i , V i , D i and, as its output, a memory address.
- the U i and V i input values (8 bits each) are connected to respective U and V registers 202 and 204 for temporarily storing these values.
- the output terminals of each of the U and V registers are connected to a multiplexer (MUX) 206 and a texture mapping address generator (TMAG) circuit 208.
- MUX multiplexer
- TMAG texture mapping address generator
- the output of each register is also connected to a "shift right 1" (SR1) circuit 210 and 212 that shifts the 8 bit output one bit to the right and inserts a zero in the most significant bit location.
- SR1 shift right 1
- each of the SR1 circuits is connected to an input of its respective register such that the register can replace the present value in the register with the "shifted" value.
- One of the three inputs to each of the U and V registers are elected by a MUX (not shown) within each register block.
- each of the registers in this block diagram have an input enable port that, when asserted by a microprocessor, enables the register to be updated with a new input value.
- the input enable ports are not shown. These ports are controlled by a microprocessor or microcontroller that ensures proper timing of the register operation to accomplish the inventive method (see FIGS. 3A and 3B).
- the MUX 206 selects either the output of the U register or the output of the V register as its output.
- the MUX output forms an A input to a logical right shifter 214.
- a B input to the logical right shifter is formed by the D i value.
- the D i input value is temporarily stored in a D register 216 before being used as the B input to the logical right shifter 214.
- the logical right shifter shifts the value at its A input by the number of bits defined by the value at its B input. As such, using the logical right shifter, either the U or V register value is shifted to the right by D bits. The output of the shifter forms another input to both the U and V registers. As such, this shifted value can be selected by either register to replace the value presently stored in the register.
- the TMAG circuit 208 has as its input the output of the U and V registers. Additionally, a third value, a stride register value, is an input to this circuit.
- a select input signal on a select port 218 selects whether the stride register value is used by the TMAG circuit or whether a stride register value shifted to the right by one bit is used. In either case, the TMAG circuit uses the selected stride value and the U and V register values to generate an offset value on line 220. The specific nature of this offset value is described below with respect to step 308 of FIG. 3A.
- the offset value is passed through a MUX 222 to one of a plurality of offset registers 224. Additionally, one of the plurality of offset registers stores a value formed by concatenating the TID value with the D i value.
- the base registers 226 contain the memory addresses for the base locations of each pyramid. These base registers are filled as the pyramids are stored in memory, e.g., during system initialization.
- the base address to use is selected by MUX 230 and the particular offset register value to use is selected by MUX 232.
- MUX 230 When an address in a selected base register is combined (added together using summer 238) with a value in a selected offset register, the generator 200 produces a memory address identifying a specify pixel value location in a particular pyramid.
- the offset register value is controlled by the inventive generator 200 such that the eight memory locations needed for the interpolation process can be rapidly determined.
- the stride register 234 is initially loaded with a value equivalent to the number of pixels in a row of pixels within a pyramid level indicated by D i .
- the value of the stride register is always a power of 2, e.g., 2, 4, 8, 16, 32, and so on.
- This value is updated with either the present value of the register shifted to the right by one bit, e.g., using the right shift circuit 236, the square of the present value, the same value, the present value plus one, or the present value minus one.
- These optional replacement values are selected from value processor 238 through a MUX 240.
- the square of the present stride register value is computed using a novel approach. Specifically, to produce the square of the 8-bit stride register value, the processor 238 performs a permutation of the present stride register value with eight zeros, as follows: ##STR1##
- the present value of the stride register is combined (using summer 242) with the present value of the offset register and then used to replace the present value of the offset register. Additionally, as stated previously, the present value of the stride register is also used as an input to the TMAG circuit 208.
- a memory address is generated. With each update of the offset register, a new memory location is generated. As such, the address generator 200 sequentially generates all eight address locations.
- a routine 300 representing the process by which the address generator 200 shown in FIG. 2 produces the eight addresses is shown as a flow chart on FIGS. 3A and B. Passing once through this routine produces eight addresses, e.g., memory addresses for pixel locations X,Y; X+1,Y; X,Y+1; X+1,Y+1; A,B; A+1,B; A,B+1; and A+1,B+1. Thus, for each pixel value required in an output image, the routine is utilized once. Thereafter, the graphics system interpolates these eight values to determine a desired pixel value for the output image. To illustrate the efficiency of the inventive routine, each block in the flow chart represents the passage of a single instruction cycle of a microprocessor controlling the operation and timing of the address generator. Thus, the entire routine is accomplished in fourteen instruction cycles.
- the routine begins at step 302 wherein the values of U i , V i , D i are respectively loaded into the U, V and D registers. Simultaneously, the values of TID and D i are concatenated and stored in one of the offset registers, i.e., as a pointer to a base value of a particular level within a particular pyramid. It is assumed that the base values of each texture pyramid are stored in the base registers upon system initialization; however, a base value for a particular level is produced using this concatenation. A MUX then selects an appropriate texture pyramid base register in accordance with the TID.
- the routine reads a memory address containing a base value for a particular pyramid level, e.g., the address of a corner of the level, and a value that represents the number of pixels in a row of that pyramid level.
- the base value is then stored in a base register and the number of pixels is stored in the stride register.
- the base register value and the stride register value requested in the previous step are returned from memory and respectively stored in a selected base register and stride register.
- these values are returned from memory as a "packed" value, i.e., both values are contained within a single word.
- the value of the V register is shifted to the right by D bits.
- the offset register (OR) is updated with a new value (a linear offset from the base value of the level) based upon a "TMAG" function of U, V and the stride register (SR) value.
- the TMAG function combines the U and V values, two 8-bit words, depending upon the present value in the stride register. Specifically, if U is assumed to have the form ABCDEFGH and V is assumed to have the form QRSTUVWX, U and V are combined as shown in Table 1 below.
- the output of the TMAG circuit is a concatenation of the U and V values in a manner that is defined by the stride register value.
- the TMAG circuit output is stored in the offset register replacing any prior value stored therein.
- the routine reads the memory location at an address defined by the base register (BR) value plus the offset register (OR) value. This address is the location of the pixel value at location X, Y (see FIG. 1). Simultaneously, the offset register value is incremented by 1 and the U value is right shifted (U>>1), e.g., the bits of the U value are shifted to the right one bit, the least significant bit is dropped and the most significant bit becomes a zero.
- BR base register
- OR offset register
- the routine receives the pixel value at location X,Y that was requested in the previous step.
- the pixel value would be stored in a register or memory location for subsequent use in an interpolation function.
- the routine reads the next pixel value located at the address indicated by the base register value plus the offset register value (BR+OR). This is the pixel located at location X+1,Y.
- the offset register is again updated, this time with the stride register (SR) value and the V value is right shifted (V>>1).
- the routine receives the pixel value at location X+1,Y that was previously read, reads the pixel value at the location defined by the base register value plus the offset register value, and decrements the offset register value by 1.
- the routine receives the pixel value at the location X+1,Y+1, reads the pixel value at the location defined by the base register value plus the offset register value, and updates the offset register with a value defined by the TMAG circuit.
- the TMAG function i.e., the concatenation of the B low bits of the U and V value, is accomplished using the right shifted stride value (SR>>1) to define the variable B.
- the routine receives the pixel value located at location X,Y+1, updates the offset register value with a combination of the previous offset register value plus the square of the stride register value, and right shifts the stride register value.
- the square of the stride register value defines a location that is one pyramid level above the previously used pyramid level.
- all four pixel values (X,Y; X+1,Y; X+1, Y+1; X,Y+1) in the D i level of FIG. 1 have been received from memory.
- the new offset register value facilitates reading the four pixel values in the next higher pyramid level, level D i+1 .
- the routine reads the pixel value at an address defined by the base register value plus the offset register value. Simultaneously, the routine updates the offset register value by incrementing the present value by 1.
- the routine receives the pixel value at the location A,B in level D i+1 that was read in the previous step, reads a new pixel value at an address defined by the base register value plus the offset register value, and updates the offset register value with the present offset register value plus the stride register value.
- the routine receives the pixel value at location A+1,B that was read during the previous step, reads a new pixel value at an address defined by the base register value plus the offset register value, and updates the offset register value by decrementing the present value by 1.
- the routine receives the pixel value at location A+1,B+1 that was read during the previous step and reads the pixel value at the address defined by the base register value plus the offset register value.
- the routine receives the pixel value at location A,B+1. At this point, the routine has addressed and received all eight pixel values necessary to perform a tri-linear interpolation and produce an appropriate target pixel value. Consequently, the routine ends at step 330.
- a graphics system e.g., a microprocessor based system controlling the inventive apparatus
- D is 2.25
- U is 9.125
- V is 15.75
- TID is 5.
- the initial base and stride values are received by the address generator as a packed 32-bit word that indicates the corner of a selected pyramid level (base) and the length of a row in that level (stride).
- base a selected pyramid level
- stride values are [1000,8], for level 2 of the pyramid.
- Level 2 of the pyramid is stored in memory addresses 1000 through 1063 and level 3 of the pyramid is stored in memory addresses 1064 through 1079.
- the numerals within the tables indicate the order in which the pixel values are determined by the routine.
- the dashes represent arbitrary pixel values that are not of concern to the present invention.
- 010 2 42) are loaded into an offset register.
- the initial base and offset register values e.g., a texture table base value represented by TID and the offset (42)
- the memory returns, within a single packed data word, the pyramid level base value (1000) and a stride register value (8).
- the memory returns pixel value 3
- the memory returns pixel value 7 and the next memory location, at address 1073, is read.
- the memory returns pixel value 8. At this time, all eight pixel values from memory locations 1026, 1027, 1034, 1035, 1069, 1070, 1073, and 1074 have been addressed and recalled from memory using fourteen instruction cycles of a microprocessor that either controls the apparatus of FIG. 2 or has the apparatus contained therein.
- the disclosed method and apparatus may find applicability in video servers, medical imaging, special effects and animation and location based entertainment systems among other applications.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Generation (AREA)
Abstract
Description
TABLE 1 __________________________________________________________________________ TMAG circuit function table U V SR value TMAG Output __________________________________________________________________________ ABCDEFGH QRSTUVWX 256 ABCDEFGHQRSTUVWX (10000000) ABCDEFGH QRSTUVWX 128 00BCDEFGHRSTUVWX (01000000) ABCDEFGH QRSTUVWX 64 (00100000) 0000CDEFGHSTUVWX ABCDEFGH QRSTUVWX 32 (00010000) 000000DEFGHTUVWX ABCDEFGH QRSTUVWX 16 (00001000) 00000000EFGHUVWX ABCDEFGH QRSTUVWX 8 (00000100) 0000000000FGHVWX ABCDEFGH QRSTUVWX 4 (00000010) 000000000000GHWX ABCDEFGH QRSTUVWX 2 (00000001) 00000000000000HX __________________________________________________________________________
TABLE 2 ______________________________________ADDRESS 0 1 2 3 4 5 6 7 ______________________________________ 1000 -- -- -- -- -- -- -- -- 1008 -- -- -- -- -- -- -- -- 1016 -- -- -- -- -- -- -- -- 1024 -- -- 1 2 -- -- -- -- 1032 -- -- 3 4 -- -- -- -- 1040 -- -- -- -- -- -- -- -- 1048 -- -- -- -- -- -- -- -- 1056 -- -- -- -- -- -- -- -- ______________________________________
TABLE 3 ______________________________________ADDRESS 0 1 2 3 ______________________________________ 1064 -- -- -- -- 1068 -- 5 6 -- 1072 -- 8 7 -- 1076 -- -- -- -- ______________________________________
Claims (14)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US08/282,296 US5596687A (en) | 1994-07-29 | 1994-07-29 | Apparatus and method for addressing pixel values within an image pyramid using a recursive technique |
PCT/US1995/009137 WO1996004613A1 (en) | 1994-07-29 | 1995-07-28 | Apparatus and method for addressing elements of an image pyramid |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US08/282,296 US5596687A (en) | 1994-07-29 | 1994-07-29 | Apparatus and method for addressing pixel values within an image pyramid using a recursive technique |
Publications (1)
Publication Number | Publication Date |
---|---|
US5596687A true US5596687A (en) | 1997-01-21 |
Family
ID=23080858
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US08/282,296 Expired - Lifetime US5596687A (en) | 1994-07-29 | 1994-07-29 | Apparatus and method for addressing pixel values within an image pyramid using a recursive technique |
Country Status (2)
Country | Link |
---|---|
US (1) | US5596687A (en) |
WO (1) | WO1996004613A1 (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5861888A (en) * | 1996-11-27 | 1999-01-19 | Vlsi Technology Inc | Method and a system for the nonlinear storage of a texture map within a linear memory device |
US5877770A (en) * | 1995-05-24 | 1999-03-02 | Sharp Kabushiki Kaisha | Texture pattern memory circuit for providing plural texel data in response to a single access operation |
US6016152A (en) * | 1997-08-01 | 2000-01-18 | Avid Technology, Inc. | Apparatus and method for non-uniform image scaling |
US6084590A (en) * | 1997-04-07 | 2000-07-04 | Synapix, Inc. | Media production with correlation of image stream and abstract objects in a three-dimensional virtual stage |
US6124864A (en) * | 1997-04-07 | 2000-09-26 | Synapix, Inc. | Adaptive modeling and segmentation of visual image streams |
US6160907A (en) * | 1997-04-07 | 2000-12-12 | Synapix, Inc. | Iterative three-dimensional process for creating finished media content |
US6184888B1 (en) * | 1997-10-31 | 2001-02-06 | Hewlett-Packard Company | Method and apparatus for rapidly rendering and image in response to three-dimensional graphics data in a data rate limited environment |
US6191793B1 (en) | 1998-04-01 | 2001-02-20 | Real 3D, Inc. | Method and apparatus for texture level of detail dithering |
US6249285B1 (en) | 1998-04-06 | 2001-06-19 | Synapix, Inc. | Computer assisted mark-up and parameterization for scene analysis |
US6266053B1 (en) | 1998-04-03 | 2001-07-24 | Synapix, Inc. | Time inheritance scene graph for representation of media content |
US6297825B1 (en) | 1998-04-06 | 2001-10-02 | Synapix, Inc. | Temporal smoothing of scene analysis data for image sequence generation |
US6522336B1 (en) * | 1997-10-31 | 2003-02-18 | Hewlett-Packard Company | Three-dimensional graphics rendering apparatus and method |
US20070013636A1 (en) * | 2005-07-14 | 2007-01-18 | Samsung Electronics Co., Ltd. | Modifying image signals for display device |
US20070060134A1 (en) * | 2005-08-09 | 2007-03-15 | Ghulam Bhatti | Structured addressing scheme for wireless networks |
US20100060646A1 (en) * | 2008-09-05 | 2010-03-11 | Apple Inc. | Arbitrary fractional pixel movement |
US11232623B2 (en) * | 2019-06-28 | 2022-01-25 | Unity IPR ApS | Method and system for creating a neural net based lossy renderer |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4992868A (en) * | 1989-02-28 | 1991-02-12 | Holtz Klaus E | True information television (TITV) and vision system |
US5097427A (en) * | 1988-07-06 | 1992-03-17 | Hewlett-Packard Company | Texture mapping for computer graphics display controller system |
US5222205A (en) * | 1990-03-16 | 1993-06-22 | Hewlett-Packard Company | Method for generating addresses to textured graphics primitives stored in rip maps |
US5224208A (en) * | 1990-03-16 | 1993-06-29 | Hewlett-Packard Company | Gradient calculation for texture mapping |
US5361386A (en) * | 1987-12-04 | 1994-11-01 | Evans & Sutherland Computer Corp. | System for polygon interpolation using instantaneous values in a variable |
-
1994
- 1994-07-29 US US08/282,296 patent/US5596687A/en not_active Expired - Lifetime
-
1995
- 1995-07-28 WO PCT/US1995/009137 patent/WO1996004613A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5361386A (en) * | 1987-12-04 | 1994-11-01 | Evans & Sutherland Computer Corp. | System for polygon interpolation using instantaneous values in a variable |
US5097427A (en) * | 1988-07-06 | 1992-03-17 | Hewlett-Packard Company | Texture mapping for computer graphics display controller system |
US4992868A (en) * | 1989-02-28 | 1991-02-12 | Holtz Klaus E | True information television (TITV) and vision system |
US5222205A (en) * | 1990-03-16 | 1993-06-22 | Hewlett-Packard Company | Method for generating addresses to textured graphics primitives stored in rip maps |
US5224208A (en) * | 1990-03-16 | 1993-06-29 | Hewlett-Packard Company | Gradient calculation for texture mapping |
Non-Patent Citations (4)
Title |
---|
Goldberg et al, "Comparative Performance of Pyramid Data Structures for Progressive Image Transmission", IEEE Transactions on Communications, vol. 39, No. 4, Apr. 4, 1991, pp. 540-548. |
Goldberg et al, Comparative Performance of Pyramid Data Structures for Progressive Image Transmission , IEEE Transactions on Communications, vol. 39, No. 4, Apr. 4, 1991, pp. 540 548. * |
Haskell, "Interpolative, Predictive, and Pyramid Transform Coding of Color Images", ICASSP '88: Acoustics, Speech & Signal Processing Conference, 1988, pp. 785-787. |
Haskell, Interpolative, Predictive, and Pyramid Transform Coding of Color Images , ICASSP 88: Acoustics, Speech & Signal Processing Conference, 1988, pp. 785 787. * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5877770A (en) * | 1995-05-24 | 1999-03-02 | Sharp Kabushiki Kaisha | Texture pattern memory circuit for providing plural texel data in response to a single access operation |
US5861888A (en) * | 1996-11-27 | 1999-01-19 | Vlsi Technology Inc | Method and a system for the nonlinear storage of a texture map within a linear memory device |
US6084590A (en) * | 1997-04-07 | 2000-07-04 | Synapix, Inc. | Media production with correlation of image stream and abstract objects in a three-dimensional virtual stage |
US6124864A (en) * | 1997-04-07 | 2000-09-26 | Synapix, Inc. | Adaptive modeling and segmentation of visual image streams |
US6160907A (en) * | 1997-04-07 | 2000-12-12 | Synapix, Inc. | Iterative three-dimensional process for creating finished media content |
US6016152A (en) * | 1997-08-01 | 2000-01-18 | Avid Technology, Inc. | Apparatus and method for non-uniform image scaling |
US6522336B1 (en) * | 1997-10-31 | 2003-02-18 | Hewlett-Packard Company | Three-dimensional graphics rendering apparatus and method |
US6184888B1 (en) * | 1997-10-31 | 2001-02-06 | Hewlett-Packard Company | Method and apparatus for rapidly rendering and image in response to three-dimensional graphics data in a data rate limited environment |
US6191793B1 (en) | 1998-04-01 | 2001-02-20 | Real 3D, Inc. | Method and apparatus for texture level of detail dithering |
US6266053B1 (en) | 1998-04-03 | 2001-07-24 | Synapix, Inc. | Time inheritance scene graph for representation of media content |
US6249285B1 (en) | 1998-04-06 | 2001-06-19 | Synapix, Inc. | Computer assisted mark-up and parameterization for scene analysis |
US6297825B1 (en) | 1998-04-06 | 2001-10-02 | Synapix, Inc. | Temporal smoothing of scene analysis data for image sequence generation |
US20070013636A1 (en) * | 2005-07-14 | 2007-01-18 | Samsung Electronics Co., Ltd. | Modifying image signals for display device |
US7839375B2 (en) * | 2005-07-14 | 2010-11-23 | Samsung Electronics Co., Ltd. | Modifying image signals for display device |
US20070060134A1 (en) * | 2005-08-09 | 2007-03-15 | Ghulam Bhatti | Structured addressing scheme for wireless networks |
US7535855B2 (en) * | 2005-08-09 | 2009-05-19 | Mitsubishi Electric Research Laboratories, Inc. | Structured addressing scheme for wireless networks |
US20100060646A1 (en) * | 2008-09-05 | 2010-03-11 | Apple Inc. | Arbitrary fractional pixel movement |
US8098262B2 (en) * | 2008-09-05 | 2012-01-17 | Apple Inc. | Arbitrary fractional pixel movement |
US11232623B2 (en) * | 2019-06-28 | 2022-01-25 | Unity IPR ApS | Method and system for creating a neural net based lossy renderer |
Also Published As
Publication number | Publication date |
---|---|
WO1996004613A1 (en) | 1996-02-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5596687A (en) | Apparatus and method for addressing pixel values within an image pyramid using a recursive technique | |
US6002407A (en) | Cache memory and method for use in generating computer graphics texture | |
US6104415A (en) | Method for accelerating minified textured cache access | |
US4808988A (en) | Digital vector generator for a graphic display system | |
US6181352B1 (en) | Graphics pipeline selectively providing multiple pixels or multiple textures | |
US4580134A (en) | Color video system using data compression and decompression | |
EP1520259B1 (en) | Method and apparatus for encoding texture information | |
US4924415A (en) | Apparatus for modifying data stored in a random access memory | |
US6057861A (en) | Mip map/rip map texture linear addressing memory organization and address generator | |
EP0272655A2 (en) | Method and apparatus for generating look-up table data | |
US7580042B2 (en) | Systems and methods for storing and fetching texture data using bank interleaving | |
EP0447229A2 (en) | Arithmetic and logic processing unit for computer graphics system | |
KR20010031608A (en) | Method and apparatus for rapidly rendering an image in response to three-dimensional graphics data in a data rate limited environment | |
KR19980019167A (en) | Image storage system and method for a block oriented image processing systems | |
US6437781B1 (en) | Computer graphics system having per pixel fog blending | |
US4241341A (en) | Apparatus for scan conversion | |
US5371840A (en) | Polygon tiling engine | |
WO1998029836A2 (en) | Circuit for determining non-homogenous second order perspective texture mapping coordinates using linear interpolation | |
CA2058585C (en) | Signal processing system including two-dimensional array transposing | |
US5657047A (en) | Method and apparatus for zooming images on a video display | |
JP3903557B2 (en) | Data conversion apparatus and image generation apparatus | |
US6084601A (en) | Corner buffer system for improved memory read efficiency during texture mapping | |
EP0112056A2 (en) | Colour video system using data compression and decompression | |
KR100261075B1 (en) | Trilinear texture mapping speed-up apparatus | |
JP2001051668A (en) | Image processor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: DAVID SARNOFF RESEARCH CENTER, INC., NEW JERSEY Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:PETERS, JOSEPH EDWARD;REEL/FRAME:007092/0717 Effective date: 19940728 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: SARNOFF CORPORATION, NEW JERSEY Free format text: MERGER;ASSIGNOR:DAVID SARNOFF RESEARCH CENTER, INC.;REEL/FRAME:009912/0460 Effective date: 19970404 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
AS | Assignment |
Owner name: DIVA SYSTEMS CORPORATION, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SARNOFF CORPORATION;REEL/FRAME:011170/0842 Effective date: 20000815 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
AS | Assignment |
Owner name: TVGATEWAY, LLC, PENNSYLVANIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:DIVA SYSTEMS CORPORATION BY HOWARD B. GROBSTEIN, CHAPTER 11 TRUSTEE;REEL/FRAME:014567/0512 Effective date: 20040421 Owner name: TVGATEWAY, LLC,PENNSYLVANIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:DIVA SYSTEMS CORPORATION BY HOWARD B. GROBSTEIN, CHAPTER 11 TRUSTEE;REEL/FRAME:014567/0512 Effective date: 20040421 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
AS | Assignment |
Owner name: SEDNA PATENT SERVICES, LLC, PENNSYLVANIA Free format text: CHANGE OF NAME;ASSIGNOR:TVGATEWAY, LLC;REEL/FRAME:015177/0980 Effective date: 20040824 Owner name: SEDNA PATENT SERVICES, LLC,PENNSYLVANIA Free format text: CHANGE OF NAME;ASSIGNOR:TVGATEWAY, LLC;REEL/FRAME:015177/0980 Effective date: 20040824 |
|
FPAY | Fee payment |
Year of fee payment: 12 |
|
AS | Assignment |
Owner name: COMCAST IP HOLDINGS I, LLC, DELAWARE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SEDNA PATENT SERVICES, LLC (F/K/A TVGATEWAY, LLC);REEL/FRAME:021570/0353 Effective date: 20080913 Owner name: COMCAST IP HOLDINGS I, LLC,DELAWARE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SEDNA PATENT SERVICES, LLC (F/K/A TVGATEWAY, LLC);REEL/FRAME:021570/0353 Effective date: 20080913 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |